【作者】 崔彬；

【导师】 童恒庆；

【作者基本信息】 武汉理工大学 ， 计算机应用技术， 2006， 硕士

【摘要】 随着数据库技术的迅速发展以及数据库管理系统的广泛应用，人们积累的数据越来越多。激增的数据背后隐藏着许多重要的信息，人们希望能够对其进行更高层次的分析，以便更好地利用这些数据。目前的数据库系统可以高效地实现数据的录入、查询、统计等功能，但无法发现数据中存在的关系和规则，无法根据现有的数据预测未来的发展趋势，缺乏挖掘数据背后隐藏的知识的手段，导致了“数据爆炸但信息贫乏”的现象。 数据挖掘技术的发展有效地满足了人们的这一愿望。因为它可以对广泛存在的大量数据进行分析，将这些数据转换成有用的信息和知识。获取的信息和知识可以广泛应用于各种应用，包括商务管理、生产控制、市场分析、工程设计和科学探索等等。数据和信息之间的鸿沟要求系统地开发数据挖掘工具，将数据坟墓转换成知识“金块”。近年来，人们在数据挖掘的理论和方法上做了大量的研究工作，并以此为基础开发出不同种类的数据挖掘工具。但是，这些工具在处理大型的多维数据集方面仍然没有取得令人满意的挖掘效果。于是，人们开始在数据挖掘中借助可视化技术，使用丰富的可视化方式将多维数据直观地表示出来，进而利用人类特有的认知能力来指导挖掘过程，最后将数据挖掘的结果以可视化的形式呈现给用户。因此，数据挖掘领域中产生了一个新的方向：可视化数据挖掘。可视化数据挖掘的目的就是使用户能够交互地浏览数据，挖掘过程等，当所要识别的不规则事物是一系列图形而不是数字表格时，人的识别速度是最快的。数据可视化与数据挖掘相辅相成，只有两者紧密结合起来才能发挥完美的作用。数据可视化主要针对数据库或数据仓库中的数据，根据数据的属性多少，可以分为一维数据可视化，二维数据可视化和多维数据可视化。广义的讲，一维和二维数据可视化技术可以看作是多维数据可视化的子集。多维性是非空间数据场的一个重要特性，所以我们在数据仓库中针对多维数据可视化进行的研究是一个很重要的课题。 多维数据可视化技术目前在国内外已经得到了广泛的研究，现在有很多常用的多维数据可视化方法，如基于几何的技术，面向图标的技术，基于层次的技术，密集象素技术等等。在本文中，我们将首先对数据挖掘技术、多维数据可视化技术、可视化数据挖掘技术进行介绍，然后我们会通过实例来介绍多维数据可视化技术在数据挖掘中的应用。更多还原

【Abstract】 Along with the rapid development of database technology and extensive application of database management system, more and more data are accumulated. Inside these quickly increased data there hide much important information, and people hope to go on higher layer analysis in order to facilitate these data better. Present database systems can realize various functions such as recording, query, statistics, and so on, but some relation and regulation among data can’t be discovered. Thus the development trend of future can’t be forecasted according to current data, and the means that mine hidden knowledge inside data is in lack, which leads to the phenomenon that data is explosive and information is scarce.The development of data mining technology can effectively satisfy the desire of people effectively with the reason that it can analyzes enormous data that abroad exist and transform them into useful information and knowledge. Information and knowledge acquired can be widely used in many fields, such as business management, production control, market analysis, engineer design and science exploring, and so on. The gap between data and information requires the systematic development for data mining tools which transform data tomb to knowledge bullion. People have done a mass of research on the theory and method of data mining recently, based on which many data mining tools have been developed. However, these tools haven’t brought about satisfactory mining effect on processing large-scale multi-dimensional data. Therefore, people begin to resort to visualization technology in data mining, which can express multi-dimensional data directly using abundant visualization means, in order to direct the process of mining, and present the outcome of data mining for users in the form of visualization. That is to say, there appears a new direction in data mining field: visual data mining. The objective of visual data mining is to make users browse data and mining process mutually. When irregular things that will be recognized are a serious of graphics but not figure tables, the recognizing speed of people is the most quick. Data visualization and data mining supplement each other, only when these two are intimately attached can intact role be exerted. Data visualization aims at data in database or data warehouse mainly, and visualization can be categorized into unilateral-dimension data visualization, planar-dimension data visualization and multi-dimension data visualization. Generally speaking,更多还原